papers

Publications (26)

cs.SD2023

Mask-CTC-based Encoder Pre-training for Streaming End-to-End Speech Recognition

Huaibo Zhao, Yosuke Higuchi, Yusuke Kida +2

Achieving high accuracy with low latency has always been a challenge in streaming end-to-end automatic speech recognition (ASR) systems. By attending to more future contexts, a str…

eess.AS2020

Recent Developments on ESPnet Toolkit Boosted by Conformer

Pengcheng Guo, Florian Boyer, Xuankai Chang +12

In this study, we present recent developments on ESPnet: End-to-End Speech Processing toolkit, which mainly involves a recently proposed architecture called Conformer, Convolution-…

eess.AS2023

BECTRA: Transducer-based End-to-End ASR with BERT-Enhanced Encoder

Yosuke Higuchi, Tetsuji Ogawa, Tetsunori Kobayashi +1

We present BERT-CTC-Transducer (BECTRA), a novel end-to-end automatic speech recognition (E2E-ASR) model formulated by the transducer with a BERT-enhanced encoder. Integrating a la…

eess.AS2021

Improved Mask-CTC for Non-Autoregressive End-to-End ASR

Yosuke Higuchi, Hirofumi Inaguma, Shinji Watanabe +2

For real-world deployment of automatic speech recognition (ASR), the system is desired to be capable of fast inference while relieving the requirement of computational resources. T…

eess.AS2020

The 2020 ESPnet update: new features, broadened applications, performance improvements, and future plans

Shinji Watanabe, Florian Boyer, Xuankai Chang +12

This paper describes the recent development of ESPnet (https://github.com/espnet/espnet), an end-to-end speech processing toolkit. This project was initiated in December 2017 to ma…

eess.AS2020

Mask CTC: Non-Autoregressive End-to-End ASR with CTC and Mask Predict

Yosuke Higuchi, Shinji Watanabe, Nanxin Chen +2

We present Mask CTC, a novel non-autoregressive end-to-end automatic speech recognition (ASR) framework, which generates a sequence by refining outputs of the connectionist tempora…

cs.LG2026

EgoBabyVLM: Benchmarking Cross-Modal Learning from Naturalistic Egocentric Video Data

Dongyan Lin, Phillip Rust, Angel Villar Corrales +19

Children acquire language grounding with remarkable robustness from limited visuo-linguistic input in ways that surpass today's best large multimodal models. Recent research sugges…

cs.CL2021

Orthros: Non-autoregressive End-to-end Speech Translation with Dual-decoder

Hirofumi Inaguma, Yosuke Higuchi, Kevin Duh +2

Fast inference speed is an important goal towards real-world deployment of speech translation (ST) systems. End-to-end (E2E) models based on the encoder-decoder architecture are mo…

eess.AS2025

Harnessing the Zero-Shot Power of Instruction-Tuned Large Language Model in End-to-End Speech Recognition

Yosuke Higuchi, Tetsuji Ogawa, Tetsunori Kobayashi

We propose to utilize an instruction-tuned large language model (LLM) for guiding the text generation process in automatic speech recognition (ASR). Modern large language models (L…

cs.CL2025

SpidR: Learning Fast and Stable Linguistic Units for Spoken Language Models Without Supervision

Maxime Poli, Mahi Luthra, Youssef Benchekroun +8

The parallel advances in language modeling and speech representation learning have raised the prospect of learning language directly from speech without textual intermediates. This…

eess.AS2021

Advancing Momentum Pseudo-Labeling with Conformer and Initialization Strategy

Yosuke Higuchi, Niko Moritz, Jonathan Le Roux +1

Pseudo-labeling (PL), a semi-supervised learning (SSL) method where a seed model performs self-training using pseudo-labels generated from untranscribed speech, has been shown to e…

eess.AS2021

A Comparative Study on Non-Autoregressive Modelings for Speech-to-Text Generation

Yosuke Higuchi, Nanxin Chen, Yuya Fujita +6

Non-autoregressive (NAR) models simultaneously generate multiple outputs in a sequence, which significantly reduces the inference speed at the cost of accuracy drop compared to aut…

cs.CL2022

A Study on the Integration of Pre-trained SSL, ASR, LM and SLU Models for Spoken Language Understanding

Yifan Peng, Siddhant Arora, Yosuke Higuchi +6

Collecting sufficient labeled data for spoken language understanding (SLU) is expensive and time-consuming. Recent studies achieved promising results by using pre-trained models in…

eess.AS2023

Segment-Level Vectorized Beam Search Based on Partially Autoregressive Inference

Masao Someki, Nicholas Eng, Yosuke Higuchi +1

Attention-based encoder-decoder models with autoregressive (AR) decoding have proven to be the dominant approach for automatic speech recognition (ASR) due to their superior accura…

eess.AS2021

Non-autoregressive End-to-end Speech Translation with Parallel Autoregressive Rescoring

Hirofumi Inaguma, Yosuke Higuchi, Kevin Duh +2

This article describes an efficient end-to-end speech translation (E2E-ST) framework based on non-autoregressive (NAR) models. End-to-end speech translation models have several adv…

eess.AS2022

Hierarchical Conditional End-to-End ASR with CTC and Multi-Granular Subword Units

Yosuke Higuchi, Keita Karube, Tetsuji Ogawa +1

In end-to-end automatic speech recognition (ASR), a model is expected to implicitly learn representations suitable for recognizing a word-level sequence. However, the huge abstract…

eess.AS2023

InterMPL: Momentum Pseudo-Labeling with Intermediate CTC Loss

Yosuke Higuchi, Tetsuji Ogawa, Tetsunori Kobayashi +1

This paper presents InterMPL, a semi-supervised learning method of end-to-end automatic speech recognition (ASR) that performs pseudo-labeling (PL) with intermediate supervision. M…

cs.SD2021

An Investigation of Enhancing CTC Model for Triggered Attention-based Streaming ASR

Huaibo Zhao, Yosuke Higuchi, Tetsuji Ogawa +1

In the present paper, an attempt is made to combine Mask-CTC and the triggered attention mechanism to construct a streaming end-to-end automatic speech recognition (ASR) system tha…

cs.CL2022

CTC Alignments Improve Autoregressive Translation

Brian Yan, Siddharth Dalmia, Yosuke Higuchi +4

Connectionist Temporal Classification (CTC) is a widely used approach for automatic speech recognition (ASR) that performs conditionally independent monotonic alignment. However fo…

eess.AS2021

Momentum Pseudo-Labeling for Semi-Supervised Speech Recognition

Yosuke Higuchi, Niko Moritz, Jonathan Le Roux +1

Pseudo-labeling (PL) has been shown to be effective in semi-supervised automatic speech recognition (ASR), where a base model is self-trained with pseudo-labels generated from unla…

cs.CL2026

SpidR-Adapt: A Universal Speech Representation Model for Few-Shot Adaptation

Mahi Luthra, Jiayi Shen, Maxime Poli +14

Human infants, with only a few hundred hours of speech exposure, acquire basic units of new languages, highlighting a striking efficiency gap compared to the data-hungry self-super…

eess.AS2023

BERT Meets CTC: New Formulation of End-to-End Speech Recognition with Pre-trained Masked Language Model

Yosuke Higuchi, Brian Yan, Siddhant Arora +3

This paper presents BERT-CTC, a novel formulation of end-to-end speech recognition that adapts BERT for connectionist temporal classification (CTC). Our formulation relaxes the con…

eess.AS2024

Predictive Speech Recognition and End-of-Utterance Detection Towards Spoken Dialog Systems

Oswald Zink, Yosuke Higuchi, Carlos Mullov +2

Effective spoken dialog systems should facilitate natural interactions with quick and rhythmic timing, mirroring human communication patterns. To reduce response times, previous ef…

eess.AS2022

ESPnet-ONNX: Bridging a Gap Between Research and Production

Masao Someki, Yosuke Higuchi, Tomoki Hayashi +1

In the field of deep learning, researchers often focus on inventing novel neural network models and improving benchmarks. In contrast, application developers are interested in maki…

eess.AS2024

End-to-End Speech Recognition with Pre-trained Masked Language Model

Yosuke Higuchi, Tetsuji Ogawa, Tetsunori Kobayashi +1

We present a novel approach to end-to-end automatic speech recognition (ASR) that utilizes pre-trained masked language models (LMs) to facilitate the extraction of linguistic infor…

eess.AS2022

Improving non-autoregressive end-to-end speech recognition with pre-trained acoustic and language models

Keqi Deng, Zehui Yang, Shinji Watanabe +3

While Transformers have achieved promising results in end-to-end (E2E) automatic speech recognition (ASR), their autoregressive (AR) structure becomes a bottleneck for speeding up…